Introduction
Expansion in the indications for cardiac implantable electronic devices (CIEDs) as well as an increase in the average life expectancy necessitate a more complex and prolonged device therapy over a patient’s lifetime. It has been reported that nearly half of the patients undergoing CIED implantation receive some form of anticoagulant or antiplatelet treatment due to multimorbidity.1,2 Perioperative management of these patients is challenging due to an increased risk of bleeding. A significant bleeding complication (SBC) is a potentially life-threatening condition. Considering the worldwide implantation rate, the medical and financial burden of postoperative SBCs also seems to be relevant.2
In clinical practice, simple scoring systems can serve as a useful tool to help providers estimate the risk of SBCs. However, to our knowledge, there is currently no bleeding score dedicated to patients undergoing CIED surgery. Therefore, there is a need for an accurate risk scoring system that would use readily available clinical information to predict the occurrence of SBCs after CIED procedures, especially in patients who use drugs affecting hemostasis. To address this need, we aimed to develop a new scoring system for predicting SBCs after CIED procedures.
Patients and methods
This was a prospective study including 1100 consecutive patients who underwent CIED (pacemaker, implantable cardioverter defibrillator [ICD], cardiac resynchronization therapy [CRT]–pacemaker [CRT-P], or CRT–defibrillator [CRT-D]) primary implantation, generator replacement, or system upgrade) in our department between June 2015 and June 2018.
Patients
The characteristics of patients, including demographic data, indications for the CIED procedure, comorbidities, and indications for and the type of chronic antiplatelet therapy (APT) or anticoagulant therapy (ACT), as well as procedural details for postimplant wound management were collected at the time of the CIED surgery. All comorbidities were diagnosed according to the current guidelines of the relevant international societies, and the appropriate treatment recorded in the documentation confirmed the diagnosis of a specific diseases. However, it should be emphasized that uncontrolled arterial hypertension was diagnosed when systolic blood pressure was at least 160 mm Hg and/or diastolic blood pressure was at least 100 mm Hg on 2 or more independent measurements during hospitalization. Patients were divided into 4 main groups: 1) ACT group (vitamin K antagonist [VKA] in doses dependent on the international normalized ratio [INR] level, direct oral anticoagulant [DOAC], or low-molecular-weight heparin [LMWH]); 2) APT group (single antiplatelet therapy [SAPT] with aspirin or clopidogrel at a dose of 75 mg/d or double antiplatelet therapy [DAPT] with aspirin [75 mg/d] and clopidogrel [75 mg/d], or aspirin [75 mg/d] and ticagrelor [180 mg/d], or aspirin [75 mg/d] and prasugrel [10 mg/d]); 3) triple antithrombotic therapy (TAT) group (DAPT with VKA or DOAC); and 4) control group (patients not receiving ACT or APT). Moreover, the ACT group was divided into 4 subgroups according to the type of an anticoagulant: 1) VKA I (warfarin or acenocoumarol) for patients with an INR of 2 or higher on the day of the procedure; 2) VKA II for patients with an INR of less than 2 on the day of the procedure; 3) DOAC (rivaroxaban, apixaban, or dabigatran in renal dosing [edoxaban is not registered in our country]); and 4) LMWH (enoxaparin / dalteparin or nadroparin in weight-adjusted dosing) (Figure 1). An antiplatelet agent was administered during the procedure. Anticoagulant agents were managed according to European Heart Rhythm Association [EHRA] recommendations. In patients with an intermediate and high risk of a thromboembolic event, VKA was administered perioperatively in patients with an INR of 3 or less (or 3.5 or less in patients with mechanical valve prosthesis), but in patients with a low thromboembolic risk, it was discontinued 3 to 4 days prior to the procedure.3 Treatment with DOAC was interrupted for 24 to 36 hours during the procedure, the duration depending on the drug type and renal function. Of note, patients in the LMWH subgroup were on chronic heparin treatment, and there were no patients on a bridging therapy with LMWH.
Figure 1. Study groups depending on the use of periprocedural therapy affecting hemostasis
Abbreviations: ACT, anticoagulant therapy; APT, antiplatelet therapy; DAPT, double antiplatelet therapy; DOAC, direct oral anticoagulant; INR, international normalized ratio; LMWH, low-molecular-weight heparin; SAPT, single antiplatelet therapy; TAT, triple antithrombotic therapy; VKA, vitamin K antagonist; VKA I, vitamin K antagonist with INR ≥2 on the day of procedure; VKA II, vitamin K antagonist with INR <2 on the day of procedure
Written informed consent was obtained from each patient included in the study. The patients’ privacy was protected by the anonymization of all data. Exclusion criteria were pregnancy, refusal to sign the informed consent form, lead and / or device extraction surgery, diagnosed thrombophilia or thrombocytopenia, and inability to attend follow-up visits for logistic reasons. The study protocol (Supplementary material) was in accordance with the Declaration of Helsinki and was approved by the bioethics committee of the Poznan University of Medical Sciences in Poznań, Poland (no. 613/15).
Significant bleeding complications
All patients were evaluated for SBCs occurring during hospitalization or during a 30-day follow-up. The complications included significant pocket hematoma (SPH) or significant bleeding (SB). Significant pocket hematoma was defined as a swelling and painful mass extending the margin of the generator, requiring surgical intervention and / or prolonging hospitalization for at least 24 hours after the CIED procedure due to interruption of ACT or APT. Significant bleeding was defined according to International Society on Thrombosis and Hemostasis criteria as fatal or symptomatic bleeding in a critical area (eg, pericardial or gastrointestinal) or any bleeding causing a decrease in hemoglobin concentrations of more than 1.24 mmol/l or leading to transfusion of 2 or more units of red blood cells.4
Statistical analysis
Patient characteristics were expressed as frequency (percentage) for categorical variables and medians (interquartile ranges) for continuous variables. None of the assessed parameters had a normal distribution as assessed using the Shapiro–Wilk test.
Categorical variables were compared using the 2-tailed Fisher exact test or the χ2 test as appropriate, and continuous variables were assessed using the Kruskal–Wallis analysis of variance test. A 2-tailed α of 0.05 was considered significant.
Univariate logistic regression was used to evaluate the magnitude of association between potential risk factors and SBCs. A multivariable analysis was performed with selected variables that were significant in the univariate analysis. A P value of less than 0.05 was considered significant. In each model, the odds ratio (OR) for each independent variable was determined with a confidence interval (CI) of 95%. A receiver operating characteristic curve analysis was performed to determine the cutoff values for the predictive level of SBC development and for the evaluation of the models. Independent predictors were then assigned weighted points. The β regression coefficients were rounded to the nearest integer to derive weights (weighing scheme, Beta/integer), and the points were summed up across the predictors for a total score for each patient. The performance of the score in terms of SBC prediction was evaluated by the receiver operating characteristic curve analysis, and its predictive ability was determined by using c-statistics as well as by calculation of the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, as well as positive and negative likelihood ratio for SBC development. All statistical analyses were performed using Statistica version 13.7 (StatSoft, Inc., Tulsa, Oklahoma, United States).
Results
The number of patients in the ACT, APT, TAT, and control groups was 384, 366, 96, and 254, respectively (Figure 1). The median patient age was 70 years (range, 18–97 years). Of the 1100 patients, 61.3% were male and 38.7% were female. Patient and procedural characteristics at baseline are presented in Table 1.
Characteristics | All patients (n = 1100) | ACT (n = 384) | APT (n = 366) | TAT (n = 96) | Controls (n = 254) | P value | ||
|---|---|---|---|---|---|---|---|---|
Demographic data | ||||||||
Age, y, median (IQR) | 70 (60–79) | 74 (64–81) | 69 (62–78) | 70 (61–78) | 63 (43–77) | <0.001 | ||
Sex | Male | 674 (61.3) | 224 (58.9) | 256 (70.3) | 71 (74.2) | 121 (47.6) | <0.001 | |
Female | 426 (38.7) | 160 (41.1) | 108 (29.7) | 25 (26) | 133 (52.1) | |||
BMI, kg/m2, median (IQR) | 27 (24.2–30.5) | 27 (24.5–31) | 27.1 (24.3–30.5) | 26.2 (23–29.4) | 26.4 (23.3–30) | 0.01 | ||
LVEF, %, median (IQR) | 50 (32–57) | 50 (35–55) | 45 (30–55) | 35 (30–55) | 55 (50–60) | <0.001 | ||
Indication for antiplatelet or anticoagulant therapy | ||||||||
Coronary artery disease | 485 (44.6) | 111 (28.9) | 273 (74.8) | 96 (100) | – | <0.001 | ||
Previous myocardial infarction | 365 (33.2) | 76 (19.8) | 215 (58.7) | 72 (75) | – | <0.001 | ||
Percutaneous coronary intervention | 386 (35.1) | 77 (20.1) | 216 (59.1) | 89 (92.7) | – | <0.001 | ||
Coronary artery bypass grafting | 103 (9.4) | 30 (7.8) | 53 (14.5) | 20 (20.8) | – | <0.001 | ||
Previous ischemic stroke | 74 (6.7) | 36 (9.4) | 24 (6.6) | 13 (13.5) | – | <0.001 | ||
Transient ischemic attack | 42 (3.8) | 30 (7.8) | 8 (2.2) | 4 (4.2) | – | <0.001 | ||
Presence of valve prosthesis | Total | 79 (7.1) | 46 (12) | 8 (2.2) | 17 (17.7) | – | <0.001 | |
Mechanical | 61 (5.5) | 40 (10.4) | – | 13 (13.5) | – | <0.001 | ||
Biological (TAVI) | 18 (1.6) | 6 (1.6) | 8 (2.2) | 4 (4.2) | – | 0.05 | ||
Atrial fibrillation / atrial flutter | 409 (37.2) | 318 (82.8) | 15 (4.1) | 76 (79.2) | – | <0.001 | ||
Previous pulmonary embolism / venous thrombosis | 28 (2.5) | 17 (4.4) | – | 11 (11.5) | – | <0.001 | ||
Antiphospholipid syndrome | 6 (0.5) | 6 (1.6) | – | – | – | <0.001 | ||
Other medical history parameters | ||||||||
Congestive HF | 603 (54.8) | 296 (77.1) | 223 (60.9) | 57 (59.4) | 27 (10.6) | <0.001 | ||
HF NYHA class I–II | 460 (41.8) | 249 (64.9) | 160 (43.7) | 29 (30.2) | 22 (8.7) | <0.001 | ||
HF NYHA class III–IV | 143 (13) | 47 (12.2) | 63 (17.2) | 28 (29.2) | 5 (2) | <0.001 | ||
Arterial hypertension | 785 (71.4) | 291 (75.8) | 289 (80) | 71 (74) | 134 (52.8) | <0.001 | ||
Renal failure (GFR <60 ml/min/m2) | 267 (24.3) | 115 (30) | 89 (24.3) | 30 (31.3) | 33 (13) | <0.001 | ||
Type 2 diabetes mellitus | 254 (23.1) | 102 (26.6) | 100 (27.3) | 19 (19.8) | 33 (13) | <0.001 | ||
Anemia | 25 (2.3) | 8 (2.1) | 6 (1.6) | 4 (4.2) | 7 (2.8) | 0.30 | ||
Malignancy | 69 (6.3) | 26 (6.8) | 26 (7.1) | 2 (2.1) | 15 (5.9) | 0.15 | ||
Previous hemorrhagic stroke | 6 (0.5) | 4 (1) | 1 (0.3) | 1 (1) | – | 0.41 | ||
Medications | ||||||||
β-Blockers | 713 (64.8) | 253 (66) | 256 (70) | 86 (89.6) | 118 (46.5) | <0.01 | ||
Calcium channel blockers | 279 (25.4) | 65 (16.9) | 58 (15.8) | 64 (66.6) | 92 (36.2) | <0.01 | ||
Sartans | 71 (6.5) | 30 (7.8) | 19 (5.2) | 9 (9.4) | 13 (5.1) | 0.05 | ||
ACEIs | 625 (56.8) | 250 (65.1) | 230 (62.8) | 90 (93.8) | 55 (21.6) | <0.001 | ||
Diuretics | 669 (60.8) | 261 (67.9) | 242 (66.1) | 48 (50) | 118 (46.5) | 0.11 | ||
Statins | 607 (55.2) | 200 (52.1) | 300 (81.9) | 96 (100) | 11 (4.3) | <0.001 | ||
Antiarrhythmic drugs | 238 (21.6) | 157 (40.9) | 3 (0.8) | 76 (79.2) | 2 (0.8) | <0.001 | ||
Others | 467 (42.5) | 170 (44.3) | 172 (47) | 40 (41.7) | 85 (33.4) | 0.09 | ||
Type of device | ||||||||
Pacemaker | 607 (55.2) | 246 (64.1) | 162 (44.3) | 38 (39.6) | 161 (63.4) | <0.001 | ||
DDD | 477 (43.3) | 154 (40.1) | 125 (34.1) | 28 (29.2) | 170 (67.2) | <0.001 | ||
VVI | 108 (9.8) | 88 (29.2) | 10 (2.7) | 10 (10.4) | – | <0.001 | ||
AAI | 22 (2) | 4 (1) | 9 (2.5) | – | 9 (3.6) | 0.08 | ||
CRT | 214 (19.5) | 75 (19.5) | 94 (25.6) | 27 (28.1) | 18 (7.1) | 0.06 | ||
CRT-D | 196 (17.8) | 65 (16.9) | 88 (24) | 25 (26) | 18 (7.1) | <0.01 | ||
CRT-P | 18 (1.6) | 10 (2.6) | 6 (1.6) | 2 (2.1) | – | 0.3 | ||
ICD | 279 (25.4) | 63 (16.4) | 129 (35.6) | 31 (32.3) | 56 (22) | <0.001 | ||
DDD | 95 (8.6) | 19 (4.9) | 36 (9.8) | 9 (9.4) | 31 (12.2) | 0.01 | ||
VVI | 184 (16.8) | 44 (11.5) | 93 (25.4) | 22 (22.9) | 25 (9.9) | <0.01 | ||
Type of procedure | ||||||||
Primary implantation | 516 (46.9) | 150 (39.1) | 184 (50.3) | 66 (68.7) | 114 (44.9) | <0.001 | ||
Generator replacement | 472 (42.9) | 184 (47.9) | 152 (41.5) | 21 (21.9) | 115 (45.3) | <0.001 | ||
System upgrade | 112 (10.2) | 50 (13) | 29 (7.9) | 9 (9.4) | 24 (9.5) | 0.001 | ||
Type of wound suturing | ||||||||
Continuous intradermal absorbable | 662 (69.3) | 288 (75) | 283 (77.1) | 70 (76.9) | 121 (47.6) | <0.001 | ||
Nonabsorbable single | 338 (30.7) | 95 (25) | 84 (22.9) | 26 (27.1) | 133 (52.4) | <0.001 | ||
Procedure duration, min, median (IQR) | ||||||||
Primary implantation | 38 (16–63) | 38 (16–63) | 35 (18–45) | 39 (25–58) | 39 (22–53) | 0.15 | ||
Generator replacement | 16 (10–21) | 16 (10–21) | 17 (10–21) | 16 (14–21) | 15 (10–21) | 0.12 | ||
System upgrade | 35 (18–60) | 37 (18–60) | 38 (16–59) | 30 (20–54) | 32 (23–55) | 0.1 | ||
Data are presented as number (percentage) unless otherwise indicated. Abbreviations: AAI, atrial demand pacing; ACEI, angiotensin-converting enzyme inhibitor; BMI, body mass index; CHF, congestive heart failure; CRT, cardiac resynchronization therapy; CRT-D, cardiac resynchronization therapy with defibrillator; CRT-P, cardiac resynchronization therapy with pacing; DDD, atrial and / or ventricular demand pacing; GFR, glomerular filtration rate; HF, heart failure; ICD, implantable cardioverter defibrillator; IQR, interquartile range; LBBB, left bundle branch block; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; TAVI, transcatheter aortic valve implantation; VVI, ventricular demand pacing; others, see Figure 1 | ||||||||
The incidence of 30-day CIED surgery–related SBCs in the entire study population was 4.5% (50 of the 1100 patients). Most frequently, SBCs occurred in the TAT group (8.3%), followed by the ACT (6.5%) and APT (3.8%) groups. In the ACT subgroup analysis, SBCs were most frequent in the LMWH subgroup (22.2%; 8 of 36 patients). Data on the frequency of SBCs are shown in Table 2.
Bleeding complications | All patients (n = 1100) | ACT subgroups | APT (n = 366) | TAT (n = 96) | Controls (n = 254) | ||||
|---|---|---|---|---|---|---|---|---|---|
All ACT (n = 384) | VKA I (n = 100) | VKA II (n = 112) | DOAC (n = 136) | LMWH (n = 36) | |||||
All SBCs | 50 (4.5) | 25 (6.5) | 4 (4) | 7 (6.3) | 6 (4.4) | 8 (22.2) | 14 (3.8) | 8 (8.3) | 3 (1.2) |
SPH requiring evacuation | 6 (0.4) | 4 (1) | 1 (1) | 2 (1.8) | 1 (0.7) | – | 1 (0.3) | 1 (1) | – |
SPH prolonging hospitalization | 41 (3.6) | 20 (5.2) | 3 (4) | 5 (6.8) | 4 (2.9) | 8 (22.2) | 13 (3.7) | 6 (6.2) | 2 (1) |
SB | 3 (0.3) | 1 (0.3) | – | – | 1 (0.7) | – | – | 1 (1) | 1 (0.4) |
No SBCs | 1050 (95.5) | 359 (93.5) | 96 (96) | 105 (93.8) | 128 (94.1) | 28 (77.8) | 352 (96.2) | 90 (93.7) | 251 (98.8) |
Data are presented as number (percentage). Abbreviations: SB, significant bleeding; SBC, significant bleeding complication; SPH, significant pocket hematoma; others, see Figure 1 | |||||||||
Predictors of serious bleeding complications
The results of the univariate and multivariable analyses for all potential predictors are presented in Table 3. The following characteristics were identified as independent predictors of SBCs: age ≥75 years (OR, 8.1; 95% CI, 3.54–18.54); CRT/ICD surgery (OR, 5.96; 95% CI, 2.48–14.32); upgrade procedure (OR, 10.22; 95% CI, 4.05–25.78); uncontrolled arterial hypertension (OR, 4.82; 95% CI, 1.78–13.06); presence of valvular prosthesis, either biological or mechanical (OR, 7.85; 95% CI, 3.15–19.58); current malignancy (OR, 6.11; 95% CI, 1.81–20.66); coexistence of renal failure (OR, 4.28; 95% CI, 1.86–9.87); and the use of antiplatelet drugs (clopidogrel [OR, 6.69; 95% CI, 2.48–18.04] or ticagrelor [OR, 22.25; 95% CI, 4.56–108.46]). It should be noted that the use of ticagrelor was the strongest risk factor.
Predictor | Univariate analysis | Multivariable analysis | ||||
|---|---|---|---|---|---|---|
OR | 95 CI | P value | OR | 95 CI | P value | |
Age ≥75 y | 8.23 | 3.8–13.77 | <0.001 | 8.1 | 3.54–18.54 | <0.001 |
CRT / ICD surgery | 5.37 | 3.37–8.02 | <0.001 | 5.96 | 2.48–14.32 | <0.001 |
System upgrade | 5.74 | 3.16–10.43 | <0.001 | 10.22 | 4.05–25.78 | <0.001 |
Congestive heart failure | 4.18 | 2.35–7.43 | <0.001 | – | ||
Uncontrolled hypertension (BP ≥160/100 mm Hg) | 8.62 | 3.65–20.38 | <0.001 | 4.82 | 1.78–13.06 | 0.002 |
Prosthesis (biological / mechanical valvular) | 17.2 | 9.3–31.94 | <0.001 | 7.85 | 3.15–19.58 | <0.001 |
Stroke | 3.05 | 1.62–5.72 | <0.001 | – | ||
Atrial fibrillation / flutter | 2.02 | 1.15–3.57 | 0.014 | – | ||
Cancer | 2.96 | 1.33–6.56 | 0.007 | 6.11 | 1.81–20.66 | 0.004 |
Anemia | 2.68 | 1.89–5.9 | <0.001 | – | ||
Renal failure (GFR <60 ml/min/m2) | 9.9 | 5.51–17.81 | <0.001 | 4.28 | 1.86–9.87 | <0.001 |
Clopidogrel | 4.5 | 2.44–8.34 | <0.001 | 6.69 | 2.48–18.04 | <0.001 |
Ticagrelor | 9.3 | 3.16–27.57 | <0.001 | 22.25 | 4.56–108.46 | <0.001 |
DAPT | 2.82 | 1.59–5.01 | <0.001 | – | ||
Acenocoumarol | 0.96 | 0.68–1.34 | 0.8 | – | ||
Warfarin | 1.39 | 0.54–3.62 | 0.49 | – | ||
Rivaroxaban | 0.89 | 0.94–2.4 | 0.88 | – | ||
Dabigatran | 1.79 | 0.53–6.03 | 0.35 | – | ||
Apixaban | 2.62 | 0.32–21.36 | 0.37 | – | ||
LMWH | 2.82 | 1.59–5.01 | <0.001 | – | ||
Abbreviations: BP, blood pressure; CRT, cardiac resynchronization therapy; ICD, implantable cardioverter defibrillator; OR, odds ratio; others, see Figure 1 and Table 1 | ||||||
The new PACE DRAP scoring system
Based on the SBC risk factors identified in the multivariable analysis, we propose a new simple score termed PACE DRAP. PACE DRAP is an acronym corresponding to each predictor: presence of valvular prosthesis (P), uncontrolled arterial hypertension (A); cancer (C); elderly (E); device type (D); renal failure (R); antiplatelets (A); and procedure type (P). The score was created using the weighted points proportional to the β regression coefficient rounded to the nearest integer in the multivariable analysis. The maximal score is 16 points. A detailed description of the factors that are included in the PACE DRAP acronym and the appropriate scoring are presented in Table 4. The PACE DRAP score significantly predicted the occurrence of SBCs (area under the curve [AUC], 0.95; P <0.001), and a score of 6 was identified as the cutoff point for high risk of SBCs with a sensitivity of 88.24% and a specificity of 87.23% (Figure 2).
Letter designation | Risk factor | Definition | OR | β regression coefficient | Points |
|---|---|---|---|---|---|
P | Prosthesis | Biological / mechanical valvular prosthesis | 7.85 | 2.09 | +2 |
A | Arterial hypertension uncontrolled | Blood pressure ≥160/100 mm Hg | 4.82 | 1.57 | +2 |
C | Cancer | Any malignancy diagnosed or treated within the last 5 years | 6.11 | 1.81 | +2 |
E | Elderly | Age ≥75 y | 8.10 | 2.09 | +2 |
D | Device type | CRT / ICD surgery | 5.96 | 1.78 | +2 |
R | Renal failure | GFR <60 ml/min/m2 | 4.28 | 1.45 | +1 |
A | Antiplatelets | Clopidogrel | 6.69 | 1.9 | +2 |
Ticagrelor | 22.25 | 3.1 | +3 | ||
P | Procedure type | System upgrade | 10.22 | 2.32 | +2 |
Abbreviations: see Tables 1 and 3 | |||||
Figure 2. Receiver operating characteristic curve analysis of the PACE DRAP score for the prediction of significant bleeding complications within 30 days after cardiac implantable electronic device surgery: A – a high value of area under curve for the PACE DRAP score; B – high sensitivity and specificity for the cutoff point for the PACE DRAP score
Abbreviations: AUC, area under the curve
Prognostic performance of the PACE DRAP score
Our analyses showed that the final PACE DRAP model had a discriminatory ability with an accuracy of 87.26%. Among the entire studied population, 178 patients were classified as having a high risk of SBCs, 44 of whom actually developed SBCs within 30 days after CIED surgery (PPV, 24.7%; NPV, 99.3%). The detailed predictive performance of the PACE DRAP score is presented in Table 5.
Parameter | High risk (≥6 points) of SBCs |
|---|---|
Patients, n | 178 |
Serious bleeding complications, n | 44 |
Relative risk | 37.94 (95% CI, 15.92–90.38) |
Positive predictive value, % | 24.7 |
Negative predictive value, % | 99.3 |
Sensitivity, % | 88 |
Specificity, % | 87.22 |
Accuracy, % | 87.26 |
Positive likelihood ratio | 6.89 |
Negative likelihood ratio | 0.14 |
Youden index | 0.75 |
Abbreviations: see Table 2 | |
Discussion
Bleeding complications and periprocedural management of patients taking drugs that affect hemostasis constitute an increasingly significant challenge in modern cardiology, especially in the face of the growing number of CIED procedures.5,6
Significant bleeding complications
In this study, the frequency of SBCs in the entire population was quite low (4.5%). The reported incidence of SBCs varies widely, from 0.7% to 5.7%, with an increasing trend in recent years, which is probably associated with the high number of CIED procedures and an increasing percentage of more complex systems.1,7,8 Another important reason may be patients’ longevity, which results in more frequent interventions in the elderly population, who are burdened with severe comorbidities, which may increase the risk of complications.
In this study, SBCs were most frequent in the TAT group. Patients in this group underwent CIED implantation within 2 months after an acute coronary incident or drug-eluting stent implantation. Data on bleeding complications in patients receiving TAT are limited. However, other studies have shown an increased risk of bleeding, with no difference in the risk of a thromboembolic or cardiovascular event in patients receiving TAT compared with those receiving warfarin and clopidogrel.9,10
Previous studies showed that periprocedural bridging therapy significantly increases the risk of pocket hematoma.8 One of the reasons for the increased incidence of SPHs in patients on LMWH seems to be postoperative use of heparin.11 Although we demonstrated quite a high incidence of SPHs in the LMWH group (22%; 8 of 36 patients), the administration of heparin was not a significant risk factor in the multivariable analysis, probably due to a small number of patients receiving this form of ACT. According to current EHRA guidelines, the use of LMWH in the periprocedural period should be avoided in patients undergoing CIED surgery.3
The PACE DRAP score for risk assessment of significant bleeding complications
Presence of valvular prosthesis (P)
The prosthetic valve is an important risk factor of SBCs. In this study, we included patients with mechanical prosthesis as well as those with biological prosthesis. Previously, Ishibashi et al12 found an over 7-fold increase in the risk of SPH in patients with valvular disease compared with those undergoing electrotherapy procedures. Similar results were reported in the ROCKET AF study, where patients with severe valvular disease had a significantly higher bleeding rate regardless of the type of an anticoagulant used and other factors.13 In the present study, no CIEDs had been implanted in patients with severe valvular disease prior to surgical treatment. A possible cause of an increased bleeding rate in patients with severe valvular disease or valvular prosthesis may be turbulent blood flow through the prosthesis or the affected valve, which affects the morphotic components of blood, especially platelets.
Arterial hypertension uncontrolled (A)
In this study, the presence of uncontrolled arterial hypertension was associated with a nearly 5-fold increase in the risk of SBCs. Previous studies did not show a relationship between arterial hypertension and the incidence of bleeding complications after CIED surgery.8,12 However, it was documented that arterial hypertension increases the risk of bleeding in patients using VKA independently of the INR level.14
Arterial hypertension represents a group of factors that may be modified by adequate pharmacotherapy and patient’s education.
Cancer (C)
The presence of a malignancy was associated with a 6-fold higher risk of SBCs. The most common types of cancer were breast (2.1%) and colorectal cancer (2%). No previously published studies included patients with malignancies. Although it is widely believed that cancer increases the risk of thromboembolic events due to hypercoagulability, previous studies have demonstrated that malignancy also increases the bleeding risk in patients on anticoagulation, even 2.5- to 6-fold as compared with patients without cancer, and is correlected with the cancer stage.15 Potential causes include prior radiotherapy, chemotherapy, or immunotherapy, and the use of nonsteroidal anti-inflammatory drugs in pain therapy. Patients may also be predisposed to bleeding because of targeted treatment–induced platelet dysfunction. However, chemotherapy-induced thrombocytopenia was not found to affect the occurrence of bleeding.16
Elderly (age ≥75 years)
In this study, the risk of SBCs was significantly higher in patients aged 75 years or older. One of the potential reasons may be venous anatomy, especially vessel tortuosity, which significantly hinders venous access and may increase the number of attempts to puncture the vessel.17 Additionally, Armaganijan et al18 showed that older people are at increased risk of pneumothorax and intracardiac electrode dislocation. Implanting the electrodes by puncture of the axillary or elbow vein using ultrasound or angiographic control may reduce the frequency of SBCs in this group.18
Device type (D)
We showed that implantation of a CRT-P or ICD increased the risk of SBCs almost 6-fold. This is in line with a study by Yang et al,19 who found that ICD or CRT-P implantation increased the risk of bleeding by 36% as compared with pacemaker implantation. This can be explained by the more complex structure, larger size, and increased stiffness of the ICD and CRT electrodes.
Renal failure (R)
In our study, renal failure (estimated glomerular filtration rate [eGFR] <60 ml/min/1.73 m2) increased the risk of SBCs more than 4-fold. Similar results have been reported in the literature.20 Tompkins et al7 showed that an eGFR of 15 to 29 ml/min/1.73 m² increases the risk of CIED-related SBCs more than 3-fold, and end-stage renal disease (eGFR <15 ml/min/1.73 m²), even 7-fold. Buiten et al21 proved that an eGFR of less than 30 ml/min/1.73 m² almost triples the risk of SPH. Platelet dysfunction and altered interactions, vascular endothelium, reduced nitric oxide synthesis, together with anticoagulation and fibrinolysis abnormalities, as well as malnutrition, appear to be responsible for the increased bleeding rate.22
Antiplatelets (A)
Antiplatelet strategy is a strong predictor of bleeding after CIED surgery. In this study, DAPT with clopidogrel increased the risk of SBCs almost 7-fold, while the use of ticagrelor increased it more than 21 times. Aspirin monotherapy did not significantly affect bleeding. Similar results have been published before.19,23 In our work, the use of ticagrelor was the strongest risk factor for SBCs, which is why it is assigned 3 points on the PACE DRAP scale. It seems that the differences in the bleeding rate between clopidogrel and ticagrelor are due to the difference in the effectiveness in inhibiting platelet activity between these two P2Y12 receptor inhibitors and their metabolism. As ticagrelor is an active drug, which means that it does not need to be converted to an active form, it acts faster than clopidogrel. Moreover, the resistance to ticagrelor is nonsignificant. To our knowledge, this is the first study to assess the effect of ticagrelor on SBCs in patients undergoing any type of a CIED procedure, including CRT.
Procedure type (P)
We demonstrated that system upgrades increase the risk of SBCs 10-fold. These results are consistent with previous literature data.24 This can be explained by the additional dissection of fibrous tissue around the device during subsequent procedures, and greater manipulation of surrounding structures may cause damage of local blood vessels.22 Furthermore, implanting a new electrode may contribute to direct injury and disruption of the vein or heart wall, resulting in local bleeding or, in the event of major injury, tamponade. Our results, in combination with previous data, confirm that all types of reintervention are associated with a greater risk of complications compared with first-time CIED surgery.
The simple PACE DRAP scoring system could be used as a quick and practical tool for SBCs risk estimation but requires external validation. Our internal validation showed that PACE DRAP was able to discriminate patients at high risk of SBCs with high accuracy, sensitivity, and specificity. However, other risk factors, such as the use of nonprescription drugs, should also be considered. It should be emphasized that several of the identified risk factors, such as hypertension, renal failure stage, or the use of antiplatelets, are potentially modifiable.
Study limitations
This study has several limitations, the main being its observational design, which prevented an equitable division of the analyzed group in terms of the type of APT and ACT used. The study included a small number of patients from a single center. Another important limitation is the heterogeneity of the analyzed group in terms of demographic and clinical characteristics, which could potentially affect the results. Although all the prescription drugs that affect the hemostatic system were included in our data set, the use of nonprescription drugs would have gone undetected. Bleeding episodes or other significant comorbidities only revealed in the primary care setting might not have been visible in our data set and may thus be underrepresented. Another limitation is that we could not validate the accuracy of the PACE DRAP scoring system using other cohorts.
Conclusions
We identified strong predictors of SBCs, several of which are potentially modifiable. We devised PACE DRAP, a simple bedside scoring system that may enable the identification of patients at high risk of SBCs related to CIED surgery without the need for laboratory testing.
Sylwia Sławek-Szmyt, MD, PhD, 1st Department of Cardiology,
December 4, 2019.
January 31, 2020.
February 6, 2020.
SS and AA conceived the concept of the study, contributed to the study design, analyzed the data, interpreted the results, and wrote the paper. MG, KS, WS, MW, TS, LC-M, ML, and PM contributed to data acquisition, analysis, and interpretation. All authors discussed the results and contributed to the final manuscript.
None declared.
Sławek-Szmyt S, Araszkiewicz A, Grygier M, et al. PACE DRAP: a simple score for predicting significant bleeding complications after cardiac implantable electronic device surgery. Pol Arch Intern Med. 2020: 130: 206-215. doi:10.20452/pamw.15180
- 1.
- Deharo JC, Sciaraffia E, Leclercq C, et al. Perioperative management of antithrombotic treatment during implantation or revision of cardiac implantable electronic devices: the European Snapshot Survey on Procedural Routines for Electronic Device Implantation (ESS-PREDI). Europace. 2016; 18: 778-784.Crossref
- 2.
- Raatikainen MJP, Arnar DO, Merkely B, et al. A decade of information on the use of cardiac implantable electronic devices and interventional electrophysiological procedures in the European Society of Cardiology countries: 2017 Report from the European Heart Rhythm Association. Europace. 2017; 19: ii1-ii90.Crossref
- 3.
- Sticherling C, Marin F, Birnie D, et al. Antithrombotic management in patients undergoing electrophysiological procedures. Europace. 2015; 8: 197-214.
- 4.
- Schulman S, Kearon C, Haemostasis SoCoAotSaSCotISoTa. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost. 2005; 3: 692-694.Crossref
- 5.
- Crowther MA, Cuker A. How can we reverse bleeding in patients on direct oral anticoagulants? Kardiol Pol. 2019; 77: 3-11.
- 6.
- Tripodi A, Braham S, Scimeca B, et al. How and when to measure anticoagulant effects of direct oral anticoagulants? Pol Arch Intern Med. 2018; 128; 379-385.Crossref
- 7.
- Tompkins C, McLean R, Cheng A, et al. End-stage renal disease predicts complications in pacemaker and ICD implants. J Cardiovasc Electrophysiol. 2011; 22: 1099-1104.Crossref
- 8.
- Birnie DH, Healey JS, Wells GA, et al. Pacemaker or defibrillator surgery without interruption of anticoagulation. N Engl J Med. 2013; 368: 2084-2093.Crossref
- 9.
- Dewilde WJ, Oirbans T, Verheugt FW, et al. Use of clopidogrel with or without aspirin in patients taking oral anticoagulant therapy and undergoing percutaneous coronary intervention: an open-label, randomised, controlled trial. Lancet. 2013; 381: 1107-1115.Crossref
- 10.
- Crowther MA, Eikelboom JW. Dual and triple antithrombotic therapies: current patterns of practice and controversies. Kardiol Pol. 2018; 76: 937-944.Crossref
- 11.
- Feng L, Li Y, Li J, Yu B. Oral anticoagulation continuation compared with heparin bridging therapy among high risk patients undergoing implantation of cardiac rhythm devices: a meta-analysis. Thromb Haemost. 2012; 108: 1124-1131.Crossref
- 12.
- Ishibashi K, Miyamoto K, Kamakura T, et al. Risk factors associated with bleeding after multi antithrombotic therapy during implantation of cardiac implantable electronic devices. Heart Vessels. 2017; 32: 333-340.Crossref
- 13.
- Breithardt G, Baumgartner H, Berkowitz SD, et al. Clinical characteristics and outcomes with rivaroxaban vs. warfarin in patients with non-valvular atrial fibrillation but underlying native mitral and aortic valve disease participating in the ROCKET AF trial. Eur Heart J. 2014; 35: 3377-3785.Crossref
- 14.
- Cantalapiedra A, Gutierrez O, Tortosa JI et al. Oral anticoagulant treatment: Risk factors involved in 500 intracranial hemorrhages. J Thromb Haemost. 2006; 22: 113-120.Crossref
- 15.
- Schulman S, Zondag M, Linkins L, et al. Recurrent venous thromboembolism in anticoagulated patients with cancer: management and short-term prognosis. J Thromb Haemost. 2015; 13: 1010-1118.Crossref
- 16.
- Johnstone C, Rich SE. Bleeding in cancer patients and its treatment: a review. Ann Palliat Med. 2018; 7: 265-273.Crossref
- 17.
- Nash A, Burrell CJ, Ring NJ, et al. Evaluation of an ultrasonically guided venepuncture technique for the placement of permanent pacing electrodes. Pacing Clin Electrophysiol. 1998; 21: 452-455.Crossref
- 18.
- Armaganijan LV, Toff WD, Nielsen JC, et al. Are elderly patients at increased risk of complications following pacemaker implantation? A meta-analysis of randomized trials. Pacing Clin Electrophysiol. 2012; 35: 131-134.Crossref
- 19.
- Yang X, Wang Z, Zhang Y, et al. The safety and efficacy of antithrombotic therapy in patients undergoing cardiac rhythm device implantation: a meta-analysis. Europace. 2015; 17: 1076-1084.Crossref
- 20.
- Galbusera M, Remuzzi G, Boccardo P. Treatment of bleeding in dialysis patients. Semin Dial. 2009; 22: 279-286.Crossref
- 21.
- Buiten MS, DE Bie MK, VAN DER Heijden AC, et al. Chronic kidney disease and implantable cardioverter defibrillator related complications: 16 years of experience. J Cardiovasc Electrophysiol. 2014; 25: 998-1004.Crossref
- 22.
- Aljadayel HA, AlKanj H, Koja S, et al. Does the preoperative mild renal dysfunction effect mortality and morbidity following valve cardiac surgery? Indian Heart J. 2016; 68: 138-142.Crossref
- 23.
- Douketis JD, Darvish-Kazem S, Spencer N, Tafur A. Perioperative management of patients who are receiving antiplatelet therapy: a case-based evidence-informed approach. Pol Arch Intern Med. 2018; 128: 771-778.Crossref
- 24.
- Kirkfeldt RE, Johansen JB, Nohr EA, et al. Complications after cardiac implantable electronic device implantations: an analysis of a complete, nationwide cohort in Denmark. Eur Heart J. 2014; 35: 1186-1194.Crossref